neo: Load the graph in chunks.

[nit.git] / src / neo.nit

diff --git a/src/neo.nit b/src/neo.nit
index df2792c..d661481 100644 (file)
--- a/src/neo.nit
+++ b/src/neo.nit
@@ -213,27 +213,27 @@ class NeoModel
 
        # Fill `model` using base pointed by `client`.
        fun load(model: Model): Model do
-               toolcontext.info("Locate all mentities...", 1)
-               var nodes = client.nodes_with_label(model_name)
+               var nodes: Array[NeoNode]
 
-               toolcontext.info("Preload nodes...", 1)
-               pull_all_nodes(nodes)
-               toolcontext.info("Preload edges...", 1)
-               pull_all_edges(nodes)
-
-               toolcontext.info("Build model...", 1)
+               toolcontext.info("Loading project node...", 1)
                nodes = client.nodes_with_labels([model_name, "MProject"])
                for node in nodes do to_mproject(model, node)
+               toolcontext.info("Loading groups...", 1)
                nodes = client.nodes_with_labels([model_name, "MGroup"])
                for node in nodes do to_mgroup(model, node)
+               toolcontext.info("Loading modules...", 1)
                nodes = client.nodes_with_labels([model_name, "MModule"])
                for node in nodes do to_mmodule(model, node)
+               toolcontext.info("Loading classes...", 1)
                nodes = client.nodes_with_labels([model_name, "MClass"])
                for node in nodes do to_mclass(model, node)
+               toolcontext.info("Loading class definitions...", 1)
                nodes = client.nodes_with_labels([model_name, "MClassDef"])
                for node in nodes do to_mclassdef(model, node)
+               toolcontext.info("Loading properties...", 1)
                nodes = client.nodes_with_labels([model_name, "MProperty"])
                for node in nodes do to_mproperty(model, node)
+               toolcontext.info("Loading property definitions...", 1)
                nodes = client.nodes_with_labels([model_name, "MPropDef"])
                for node in nodes do to_mpropdef(model, node)
                return model
@@ -270,52 +270,6 @@ class NeoModel
                do_batch(batch)
        end
 
-       # Load content for all `nodes` from base.
-       #
-       # Content corresponds to properties and labels that are loaded in batch mode.
-       private fun pull_all_nodes(nodes: Collection[NeoNode]) do
-               var batch = new NeoBatch(client)
-               var len = nodes.length
-               var sum = 0
-               var i = 1
-               for node in nodes do
-                       batch.load_node(node)
-                       if i == batch_max_size then
-                               do_batch(batch)
-                               sum += batch_max_size
-                               toolcontext.info(" {sum * 100 / len}% done", 1)
-                               batch = new NeoBatch(client)
-                               i = 1
-                       else
-                               i += 1
-                       end
-               end
-               do_batch(batch)
-       end
-
-       # Load all edges from base linked to `nodes`.
-       #
-       # Edges are loaded in batch mode.
-       private fun pull_all_edges(nodes: Collection[NeoNode]) do
-               var batch = new NeoBatch(client)
-               var len = nodes.length
-               var sum = 0
-               var i = 1
-               for node in nodes do
-                       batch.load_node_edges(node)
-                       if i == batch_max_size then
-                               do_batch(batch)
-                               sum += batch_max_size
-                               toolcontext.info(" {sum * 100 / len}% done", 1)
-                               batch = new NeoBatch(client)
-                               i = 1
-                       else
-                               i += 1
-                       end
-               end
-               do_batch(batch)
-       end
-
        # How many operation can be executed in one batch?
        private var batch_max_size = 1000